How Robust are LLMs to In-Context Majority Label Bias?

Open in new window